mif-induced stromal pkcβ/il8 is essential in human acute ... · microenvironment and immunology...

Microenvironment and Immunology

MIF-Induced Stromal PKCb/IL8 Is Essential inHuman Acute Myeloid LeukemiaAmina M. Abdul-Aziz1, Manar S. Shafat1, Tarang K. Mehta2, Federica Di Palma2,Matthew J. Lawes3, Stuart A. Rushworth1, and Kristian M. Bowles1,3

Abstract

Acute myeloid leukemia (AML) cells exhibit a high level ofspontaneous apoptosis when cultured in vitro but have aprolonged survival time in vivo, indicating that tissue micro-environment plays a critical role in promoting AML cell sur-vival. In vitro studies have shown that bone marrow mesen-chymal stromal cells (BM-MSC) protect AML blasts from spon-taneous and chemotherapy-induced apoptosis. Here, we reporta novel interaction between AML blasts and BM-MSCs, whichbenefits AML proliferation and survival. We initially examinedthe cytokine profile in cultured human AML compared withAML cultured with BM-MSCs and found that macrophagemigration inhibitory factor (MIF) was highly expressed by

primary AML, and that IL8 was increased in AML/BM-MSCcocultures. Recombinant MIF increased IL8 expression in BM-MSCs via its receptor CD74. Moreover, the MIF inhibitor ISO-1 inhibited AML-induced IL8 expression by BM-MSCs as wellas BM-MSC–induced AML survival. Protein kinase C b(PKCb) regulated MIF-induced IL8 in BM-MSCs. Finally,targeted IL8 shRNA inhibited BM-MSC–induced AML surviv-al. These results describe a novel, bidirectional, prosurvivalmechanism between AML blasts and BM-MSCs. Furthermore,they provide biologic rationale for therapeutic strategies inAML targeting the microenvironment, specifically MIF and IL8.Cancer Res; 77(2); 1–9. �2016 AACR.

IntroductionSurvival of patients with acute myeloid leukemia (AML) is

presently poor; two thirds of young adults and 90% of olderadults die of their disease (1). Even in patients who achieveremission with chemotherapy, relapse is common and occursfrom minimal residual disease sequestered in protectiveniches in the bone marrow microenvironment (2). According-ly, it is envisaged that improved outcomes will come fromnovel treatment strategies derived from an improved under-standing of the biology of AML within the bone marrowmicroenvironment.

AML cells exhibit a high level of spontaneous apoptosis whencultured in vitro but have a prolonged survival time in vivo,indicating that the tissue microenvironment plays a critical rolein promoting AML cell survival (3–6). Knowledge of the com-plexity of the bone marrow microenvironment is increasing,especially with respect to the bonemarrowmesenchymal stromal

cells (BM-MSC), which are considered amajor protective cell type(7). BM-MSCs generate various factors whose primary functionsare to influence tumor cell survival and homing (4, 8, 9). Theapoptotic defect in AML is not cell autonomous but highlydependent on extrinsic signals derived from their microenviron-ment. The complex cell–cell interactions between the AML tumorcells and their microenvironment are therefore essential fortumor growth and survival and thus present an attractive targetfor novel drug therapies.

Macrophage migration inhibitory factor (MIF) is a pleiotropiccytokine, which under normal conditions regulates cell-medi-ated immunity and inflammation (10). In cancer, MIF is over-expressed in a number of solid tumors, including breast, pros-tate, and colon cancers (11–13). MIF has also been shown to beoverexpressed in various blood cancers, including chronic lym-phocytic leukemia (CLL; ref. 14). In CLL, MIF is expressed by themalignant cells and induces protective IL8 release in an auto-crine-dependent manner. Blocking either MIF or IL8 reducessurvival of CLL. The increased secretion of IL8 from tumor cellsis thought to have wider significance to the tumor microenvi-ronment. Serum IL8 is known to be higher in patients with AML,myelodysplasia (MDS), and non-Hodgkin lymphoma than innormal controls, and levels of IL8 in these patients are similar tothose found in patients with multiple organ failure of nonsepticorigin (15, 16). Furthermore, leukemic blasts from the majorityof patients with AML constitutively express IL8 (17). In addi-tion, inhibition of the IL8 receptor, CXCR2, selectively inhibitsproliferation of MDS/AML cell lines and patient samples (18).Together, these studies suggest that MIF and IL8 are functionallyimportant in regulating the survival and proliferation of mul-tiple tumors, including AML.

In the current study, we investigate how AML cells programBM-MSCs via MIF to produce the survival cytokine IL8 and

1Department of Molecular Haematology, Norwich Medical School, University ofEast Anglia, Norwich, United Kingdom. 2The Genome Analysis Centre (TGAC),Colney, Norwich, United Kingdom. 3Department of Haematology, Norfolk andNorwich University Hospitals NHS Trust, Norwich, United Kingdom.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

S.A. Rushworth and K.M. Bowles are co-senior authors of this article.

Corresponding Authors: S.A. Rushworth, University of East Anglia, NorwichResearch Park, Norwich, Norfolk NR4 7TJ, United Kingdom. Phone: 075-4725-4801; Fax: 016-0359-3752; E-mail: [email protected]; and Kristian M.Bowles, [email protected].

doi: 10.1158/0008-5472.CAN-16-1095

�2016 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org OF1

Research. on August 3, 2020. © 2016 American Association for Cancercancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-1095

http://cancerres.aacrjournals.org/

characterize the signaling pathways underlying this interdepen-dent cell–cell communication.

Materials and MethodsMaterials

Anti-PKC, MAPK, and AKT antibodies were purchased fromCell Signaling Technology. Anti-CD74, anti-CXCR2, and anti-CXCR4 antibodies were purchased from Miltenyi Biotec. Allinhibitors were purchased from Tocris. The CD74 blocking anti-body was purchased from BD Biosciences. Proteome ProfilerHuman XL array and recombinant human MIF were purchasedfrom R&D Systems. MIF ELISA was purchased from BioLegend.IL8ELISAwaspurchased fromeBioscience. All other reagentswereobtained from Sigma-Aldrich

Cell cultureFor primary cell isolation, heparinized blood was collected

from volunteers, and human peripheral bloodmononuclear cellswere isolated by Histopaque (Sigma-Aldrich) density gradientcentrifugation. AML samples that comprised less than 80% blastswere purified using the CD34 Positive Selection Kit. Cell type wasconfirmed by microscopy and flow cytometry. BM-MSCs wereisolated by bone marrow aspirates from AML patients. Mononu-clear cells were collected by gradient centrifugation and plated ingrowth medium containing DMEM and 20% FBS and 1% L-glutamine. The nonadherent cells were removed after 2 days.When 60% to 80% confluent, adherent cells were trypsinized andexpanded for 3 to 5 weeks. BM-MSCs were checked for positiveexpression of CD105, CD73, and CD90 (BM-MSC markers) andthe lack of expression of CD45 and CD34 by flow cytometry. Allpatient information, including genotype and WHO classificationfor AML and genotype and phenotype of AML BM-MSCs, areincluded in Supplementary Tables S1 and S2.

RNA extraction and real-time PCRTotal RNA was extracted from 5 � 105 cells using the Nucleic

Acid PrepStation from Applied Biosystems, according to themanufacturer's instructions. Reverse transcription was performedusing the RNA PCR Core Kit (Applied Biosystems). Relative qRT-PCR used SYBR Green technology (Roche) on cDNA generatedfrom the reverse transcription of purified RNA. After preamplifi-cation (95�C for 2minutes), the PCRswere amplified for 45 cycles(95�C for 15 seconds and 60�C for 10 seconds and 72�C for 10seconds) on a 384-well LightCycler 480 (Roche). Each mRNAexpression was normalized against GAPDH mRNA expressionusing the standard curve method.

Western immunoblotting and ELISAsSDS-PAGE and Western blot analyses were performed. Briefly,

whole-cell lysates were extracted and SDS-PAGE separation wasperformed. Protein was transferred to nitrocellulose membraneand Western blot analysis performed with the indicated antiseraaccording to theirmanufacturer's guidelines. To examineMIF andIL8 secretion into media, we used LEGEND MAX Human ActiveMIF ELISA Kit (BioLegend) andHuman IL-8 ELISA Ready-SET-Go(eBioscience).

shRNA silencing of CD74, PKCb, and IL8Five Mission shRNA targeted lentivirus particles (Sigma-

Aldrich) for each target were obtained. BM-MSCs (2 � 104 cells)were infected with each lentivirus. For all gene expression experi-

ments, the cells were incubated for 72 hours posttransfectionbefore RNA extraction.

Flow cytometryFlow cytometry formeasuring AML cell numberwas performed

on the Cube 6 (Sysmex Partec). For the AML/BM-MSC cocultures,AML cell viability was measured using flow cytometry. Afterexclusion of BM-MSCs by electronic gating using forward scatter,AML cells were counted using CD34 gating.

Cytokine array expression analysisPrimary AML blasts 0.25 � 106 were cultured alone or cocul-

tured on confluent primary BM-MSCs. Conditionedmediumwasthen collected from these cultures aswell as fromBM-MSC cultureand analyzed using the Proteome Profiler Human XL CytokineArray following themanufacturer's instructions.Quantificationofcytokine optical densities was obtained with the HLimageþþsoftware (WesternVision).

Statistical analysesThe Mann–Whitney U test was used to compare test groups

where stated. Results where P < 0.05 were considered statisticallysignificant. Results represent the mean � SD of four or moreindependent experiments. We generated statistics with GraphPadPrism 5 software (GraphPad). For Western blotting, data arerepresentative images of three independent experiments.

Study approvalAML cells and BM-MSCswere obtained fromAMLpatient bone

marrowor blood following informed consent andunder approvalfrom the UK National Research Ethics Service (LRECref07/H0310/146).

ResultsBM-MSCs support AML survival

The microenvironment supports AML survival and prolifera-tion (4, 5, 19). To study the cell–cell communicationbetweenBM-MSCs and AML cells, we established a coculture system usingprimary AML cells and BM-MSCs derived from treatment-na€�veAML patients. Here, we show a significant difference in primaryAML survival when cultured on BM-MSCs for 6 and 14 dayscompared with AML blast survival when cultured in basal mediaalone (Fig. 1A; Supplementary Fig. S1). Supplementary Figure S2shows the different combinations of AML and BM-MSCs in allexperiments.

To determine what factors are responsible for improved pri-mary AML blast survival on BM-MSCs, we analyzed the profile ofcytokines and chemokines present in primary AML cultures, BM-MSC cultures, and primary AML blasts cultured in combinationwith BM-MSCs. Cytokine array profiles of the three culture con-ditions (Fig. 1B) show a consistent upregulation of IL8 in thecoculture sample media (Fig. 1C). Moreover, we also observedhigh levels ofMIF in all AML supernatants, low levels ofMIF in allBM-MSC supernatants, and high levels of MIF in the AML/BM-MSC cocultures (Fig. 1B and D). To verify these observations, wecarried out IL8- and MIF-specific ELISAs. IL8 concentrations peakat 8 and 24 hours in the AML/BM-MSC coculture supernatants(Fig. 1E), whereas MIF concentrations were high and at similarlevels in AML culture supernatants and AML/BM-MSC coculturesupernatants (Fig. 1F).

Abdul-Aziz et al.

Cancer Res; 77(2) January 15, 2017 Cancer ResearchOF2




AML-derived MIF induces IL8 expression in BM-MSCsNext, we looked to determine whether BM-MSCs needed

direct contact with AML to increase IL8 expression. RT-PCRshowed that IL8 mRNA from BM-MSCs incubated with AMLincreased by 57-fold when in direct contact and by 50-foldwhen in indirect contact with AML blasts (Fig. 2A). Thisconfirms that direct tumor cell to stromal cell contact is notnecessary for AML to induce increased IL8 expression byBM-MSCs.

Next, we examined the mRNA expression levels of MIF inprimary AML (n ¼ 5) and BM-MSC (n ¼ 5) cultures. RT-PCRshowed that primary AML cultures, but not BM-MSC cultures,express high levels of MIF mRNA under normal basal conditions(Fig. 2B). As MIF expression has been shown to be increased inAML patients compared with normal patients (20) and the abilityof MIF to induce IL8 production by primary CLL (14), wehypothesized that MIF from AML was responsible for theincreased IL8 expression in BM-MSCs. To test this hypothesis,we stimulated BM-MSCs with 100ng/mL recombinant humanMIF and assayed for IL8 mRNA and protein expression over a

period of 24 hours. We show that IL8 mRNA and proteinincreased (Fig. 2C and D) in response to MIF. To confirm thatMIF secreted from AML cells regulates IL8 expression, we usedISO-1, a nontoxic inhibitor ofMIF, which functions by binding tobioactive MIF at its N-terminal tautomerase site (21). MIF-stim-ulated BM-MSCs pretreated with ISO-1 showed a decrease inIL8 mRNA levels compared with untreated BM-MSCs (Fig. 2E).Moreover, AML survival was inhibited when cultured with BM-MSCs in the presence of ISO-1 compared with control AML–BM-MSC cultures (Fig. 2F). Together, these data confirm that MIFsecreted by AML cells induces IL8 expression in BM-MSCs.

MIF-induced IL8 upregulation is mediated through CD74Depending on the cellular context and the disease involved,

MIF signaling is mediated by its receptors CXCR2 (IL8 receptor,ILR8) and/or CXCR4 (stromal-derived factor 1 receptor),and/or CD74 (22, 23). BM-MSCs have been reported to expressall three receptors (24–26). Using CD105 as a BM-MSC markerto confirm mesenchymal cell phenotype, we show that CD74and CXCR4 are expressed but CXCR2 is not expressed on

Figure 1.

BM-MSCs support AML survival.A, AML blasts (0.25 � 106) werecocultured with primary BM-MSCs ona 12-well plate for 6 days (n¼ 20); AMLblast number was assessedusing Trypan blue exclusionhemocytometer–based counts andCD34þ staining using flow cytometry.B, Cell-free supernatants from7 individual AML patient blasts(0.25 � 106) were cocultured withprimary BM-MSCs for 24 hours and arepresentative cytokine antibodyarray of each of the cell cultureconditioned media using the HumanCytokine Proteome Profiler Array.C, Fold induction of cytokinesbetween BM-MSCs and AML culturedon BM-MSCs. Results from 7 differentprimary AML on four differentBM-MSCs. Bars, mean and SEM.Significantly upregulated cytokinesare included in the graph. D,Quantification of cytokine opticaldensity of AML only arrays(7 individual AML samples). Graphshows the top 8 cytokines expressedby AML. Bars, mean and SEM. E, IL8ELISA of each of the cell cultureconditioned media from various timepoints (three individual AML samples).F,MIF ELISA of each of the cell cultureconditioned media from various timepoints (three individual AML samples).The Mann–Whitney U test was used tocompare between treatmentgroups (� , P < 0.05).

MIF Regulates AML Microenvironment

www.aacrjournals.org Cancer Res; 77(2) January 15, 2017 OF3




all primary BM-MSCs isolated from AML patients (Fig. 3A).BM-MSCs were further characterized using CD73 and CD90,and lack of CD45 expression.

We used specific inhibitors of CXCR2, CXCR4, and CD74 todetermine which receptor/s were responsible for MIF-inducedIL8 upregulation. Inhibition of CXCR2 and CXCR4 usingpertussis toxin (a GPCR inhibitor) had no effect on MIF-induced IL8 mRNA expression (Fig. 3B). However, the anti-

CD74 blocking antibody inhibited MIF-induced IL8 expressionin BM-MSCs (Fig. 3C). These results suggest that CD74 is thedominant receptor in regulating MIF-induced IL8 expressionin AML patient-derived BM-MSCs. To further characterize thisinteraction, we used lentiviral-mediated knockdown (KD) ofCD74 in AML patient-derived BM-MSCs, confirming reducedmRNA and protein expression of CD74 after transduction withcontrol KD or CD74 KD lentivirus (Fig. 3D). Furthermore, we

Figure 2.

AML-derived MIF induces IL8expression in BM-MSCs. A, AML blastsfrom 5 patients (0.25 � 106) werecocultured with primary BM-MSCseither in direct contact (DC) or indirectcontact (IC, transwell insert) for24 hours. IL8 mRNA in BM-MSCs wasthen assessed by real-time PCR. mRNAexpression was normalized to GAPDHmRNA levels (n ¼ 5). B, BM-MSCs andprimary AML from 5 patients werecultured alone and measured for MIFmRNA levels (n¼ 5). mRNA expressionwas normalized to GAPDH mRNAlevels. C, BM-MSCs from 5 patientswere treated with rhMIF (100 ng/mL)for indicated times, and then extractedRNA was assessed for IL8 mRNA byreal-time PCR. mRNA expressionwas normalized toGAPDHmRNA levels(n ¼ 5). D, BM-MSCs from 5 patientswere treated with rhMIF (100 ng/mL)for indicated times, and then, mediawere assessed for IL8 proteinexpression by ELISA. (n ¼ 5).E, BM-MSCs from four patients werepretreated with ISO-1 (10 mmol/L) for5 minutes before treatment with rhMIF(100 ng/mL) for 4 hours and thenassessed for IL8 mRNA expression(n ¼ 4). F, BM-MSCs were pretreatedwith ISO-1 (10 mmol/L) for 5 minutesbefore the addition of primary AMLblasts from 10 patient samples for48 hours. AML blast number wasassessed using Trypan blue exclusionhemocytometer–based counts (n¼ 10).The Mann–Whitney U test was used tocompare between treatment groups(�, P < 0.05).

Abdul-Aziz et al.





demonstrate that CD74 knockdown inhibits MIF-induced IL8mRNA expression in AML patient-derived BM-MSCs (Fig. 3E).

Pharmacologic inhibition of PKCb inhibits MIF-inducedIL8 induction in BM-MSCs

We next investigated the signaling cascade in AML patient-derived BM-MSCs downstream of MIF-induced CD74 activa-tion. It has been shown that MIF binding to CD74 activatesdownstream signaling through the PI3K/protein kinase B(AKT) and MAPK signaling pathways and promotes cell pro-liferation and survival (27). In addition, Lutzny and colleaguesrecently described the activation of a PKC pathway in murinestromal cells cocultured with chronic lymphocytic leukemia(28). We treated AML-stimulated BM-MSCs with LY294002(a PI3K/Akt inhibitor), PD098059 [a MAPK kinase (MEK)1 inhibitor], or Ro-31-8220 (a PKC pan inhibitor) to determinewhich pathway/s regulate AML-induced BM-MSC IL8 mRNAinduction. We show that Ro-31-8220, the PKC inhibitor, wasable to significantly inhibit IL8 expression by approximately80%, whereas LY294002 and PD098059 had little or no effect(Fig. 4A). Similarly, we found that Ro-31-8220 was able to

inhibit IL8 expression by circa 90% in experiments where BM-MSCs were directly activated using recombinant human MIF(rhMIF; rather than AML cells; Fig. 4B). However, in addition,we observed that PD98059 was able to moderately inhibitrhMIF-induced IL8 mRNA induction in BM-MSCs by approx-imately 30% (Fig. 4B).

To clarifywhether PKC,MAPK,or both are activated in responseto MIF, we performed Western blot analysis on BM-MSCs forspecific phosphorylation of PKC isoforms, MAPK, or AKT inresponse to MIF activation. BM-MSCs were activated by AML for15 minutes or MIF treatment (100 ng/mL) for various times. Weinitially show that MIF and AML both induce phosphorylationof PKC a/bII and PKC b in BM-MSCs (Fig. 4C). BMSCs from4 patient samples treated withMIF had no increase in phosphory-lationof AKT andMAPK (Fig. 4D).Next, the PKC isoform–specificinhibitors, Go6976 (PKCa/b) and enzastaurin (PKCb), were usedto blockMIF-induced IL8 expression in BM-MSCs. Both inhibitorsshowed inhibition of MIF-induced IL8 upregulation (Fig. 4E).Finally, we used lentiviral-mediated KD for PKCb, confirmingreduced mRNA expression of PKCb after transduction of BM-MSCs with control KD or PKCb KD virus (Fig. 4F). We then

Figure 3.

MIF-induced IL8 upregulation ismediated through CD74. A, BM-MSCswere assessed for CD105, CD74,CXCR2, and CXCR4 using flowcytometry. B and C, BM-MSCs fromfour patient samples were pretreatedwith the GPCR inhibitor pertussis toxin(PTX; 100 ng/mL) or for CD74 (aCD74ab; 10mg/mL) for 30 minutes beforebeing stimulated with MIF for 4 hours.RNA was extracted and assessed forIL8 mRNA by real-time PCR. mRNAexpression was normalized to GAPDHmRNA levels (n¼4).D,BM-MSCs fromfour patient samples were infectedwith control shRNA or CD74 shRNA for72 hours and analyzed for CD74mRNAexpression by RT-PCR and proteinexpression by flow cytometry. E, BM-MSCs from four patient samples wereinfected with control shRNA or CD74shRNA for 72 hours and then treatedwith recombinantMIF and analyzed forIL8 mRNA expression by RT-PCR. TheMann–Whitney U test was used tocompare between treatment groups(� , P < 0.05).






demonstrate that knockdown of PKCb inhibits MIF-induced IL8mRNA expression (Fig. 4G). Together, these results confirm thatMIF-induced IL8 expression in AML patient-derived BM-MSCsrequires PKCb.

Targeting the MIF–PKCb–IL8 axis disrupts BM-MSC–inducedprotection of primary human AML blasts

Finally, to examine the effect of blocking IL8 on BM-MSCprotection and survival of primary AML blasts, we cocultured

Figure 4.

Inhibition of PKCb regulates AML-derived MIF-induced IL8 mRNA induction in BM-MSCs. A, BM-MSCs from four patients samples were pretreated with Ro-31-8220 (1 mmol/L), PD98059 (10 mmol/L), and LY294002 (10 mmol/L) and then incubated with primary AML blast for 4 hours. RNA was extracted andassessed for IL8 mRNA by real-time PCR. mRNA expression was normalized to GAPDH mRNA levels (n ¼ 4). B, BM-MSCs from four different sampleswere pretreated with Ro-31-8220 (250 nmol/L), PD98059 (10 mmol/L), and LY294002 (10 mmol/L) and then incubated with MIF for 4 hours. RNAwas extracted and assessed for IL8 mRNA by real-time PCR. mRNA expression was normalized to GAPDH mRNA levels (n ¼ 4). C, BM-MSCs werecultured with AML for 15 minutes or recombinant MIF (100 ng/mL) for various times. Protein was extracted and Western blotting performed. Blotswere probed for pPKCa/bII, pPKCb, PKD/PKCm, PKCd, and PKCd/q. Blots were then reprobed for b-actin to show equal sample loading. D, Fourdifferent BM-MSCs with recombinant MIF (100 ng/mL) at various times. Protein was extracted and Western blotting performed. Blots were probed for pAKTand pMAPK as well as total AKT and total MAPK. Blots were then reprobed for b-actin to show equal sample loading. E, Four different BM-MSCswere pretreated with Go 6976 (1 mmol/L) and enzastaurin (1 mmol/L) and then incubated with MIF for 4 hours. RNA was extracted and assessed forIL8 mRNA by real-time PCR. mRNA expression was normalized to GAPDH mRNA levels (n ¼ 4). F, Four different BM-MSCs were infected withcontrol shRNA or PKCb shRNA for 72 hours and analyzed for PKCb mRNA expression by RT-PCR. G, Four different BM-MSCs were infected with controlshRNA or PKCb shRNA for 72 hours, then treated with recombinant MIF and analyzed for IL8 mRNA expression by RT-PCR. The Mann–Whitney U testwas used to compare between treatment groups (� , P < 0.05).

Abdul-Aziz et al.





primary AML blasts derived from treatment-na€�ve AML patientswith BM-MSCs (either control KD or IL8 KD). First, we usedlentiviral-mediated KD for IL8. Figure 5A and B shows the mRNAexpression andprotein expression of IL8 after transduction of BM-MSCs with control KD or IL8 KD virus. Figure 5C demonstratesthat knockdown of IL8 inhibits MIF-induced IL8 mRNA expres-sion. Next, we show that KD of IL8 in BM-MSCs significantlyinhibits AML survival when in coculture compared with controlKD BM-MSCs (Fig. 5D). Finally, blocking the IL8R usingSB225002 inhibited AML survival when cultured with BM-MSCs(Fig. 5E). Taken together, these results identify a novel protumoralregulatory pathway in the AML microenvironment.

DiscussionAML is primarily a disease of the elderly with a median age at

diagnosis in the SwedishAcute LeukemiaRegistry of 72years (29).Outcomes for the 75% of patients who get AML over the age of 60remain generally poor, largely because the intensity and sideeffects of existing curative therapeutic strategies (which are com-monly used to treat younger fitter patients), coupled with patientcomorbidities, frequently limit their use in this older less fitpopulation (30). Accordingly, there is an urgent need to identifypharmacologic strategies to tackle AML, which are not onlyeffective but can also be tolerated by both older and less wellpatients. It is envisaged that treatments that target the tumormicroenvironment may well help realize this goal.

Here, we report a novel survival pathway within the humanAMLmicroenvironment, which functions as a feedback/autocrineloop involving the constitutive expression of the chemokine MIFby the AML blasts, which in turn induces IL8 expression in BM-

MSCs. Interestingly, another group showed that the repertoire ofconstitutive in vitro chemokine release from AML shows variationbetween different AML patient samples (31). We find thatalthough baseline expression of MIF by AML varies betweenpatient samples tested, all samples analyzed expressed MIF.Moreover in coculture experiments, AML patient-derived BM-MSCs were found to be ubiquitously responsive to AML-derivedMIF, which resulted in an increase in IL8 expression by the BM-MSCs. This is in keeping with similar reports on other cytokinepathways, which have shown that BM-MSCs can constitutivelyexpress various chemokines (32), and AML cells are able torespond to these chemokines (32, 33). In this study, we alsoexamined the genotype of six BM-MSCs used for the experimentsand found three of six to be normal and in other three, thegenotyping failed (Supplementary Table S2). This is apparentlyin contrast to Huang and colleagues, who found that three of fourBM-MSCs from AML patients tested had cytogenetic abnormal-ities within the stromal cells (34). Presently therefore, the inci-dence and functional consequences of cytogenetic mutationswithin stromal cells remain undefined. Nevertheless, takentogether, despite the established heterogeneity in AML, we findthe MIF/IL8 autocrine loop a constant finding across the sampleswe tested, which makes this an attractive druggable target.

IL8 is a proinflammatory chemokinewhose primary function isto activate two cell surface G protein–coupled receptors, CXCR1and CXCR2, which promote neutrophil migration and degranu-lation (35–37). Elevated IL8 secretion in tumor biology is wellcharacterized, with a number of studies showing the importanceof this chemokine in AML. In 1993, Tobler and colleaguesdescribed the constitutive expression of IL8 and its receptor inhuman myeloid leukemia, and more recently, Schinke and

Figure 5.

Targeting MIF–PKC–IL8 axis in AMLdisrupts BM-MSC–derived protection.A and B, BM-MSCs from six sampleswere infected with control shRNA orIL8 shRNA for 72 hours and analyzedfor IL8 mRNA expression by RT-PCRand protein expression by ELISA.C, BM-MSCs from six samples wereinfected with control shRNA or IL8shRNA for 72 hours and then treatedwith recombinant MIF and analyzedfor IL8 mRNA expression by RT-PCR.D, BM-MSCs were infected withcontrol shRNA or IL8 shRNA for72 hours and cocultured with AMLblasts from seven samples for 48hours. AML blast number wasassessed using Trypan blue exclusionhemocytometer–based counts(n ¼ 7). E, BM-MSCs were pretreatedwith SB225002 (100 nmol/L) for30 minutes before the addition ofprimary AML blasts from 10 samplesfor 48 hours. AML blast number wasassessed using Trypan blue exclusionhemocytometer–based counts. TheMann-Whitney U test was used tocompare between treatment groups(� , P < 0.05).






colleagues have reported that inhibition of the IL8 receptorCXCR2 selectively inhibits immature hematopoietic stem cellsfromMDS/AML samples (17, 18). In other malignancies, IL8 hasbeen characterized in endothelial cells and tumor-associatedmacrophages, suggesting that IL8 has a function in the liver andprostate tumor microenvironments (38, 39). As rodents lack adirect homologue of IL8, we purified BM-MSCs extracted frompatient bone marrow aspirates at the time of diagnosis of AML.Our study describes for the first time how AML stimulates theproduction of IL8 from BM-MSCs and inhibiting this processprevents AML survival.

Extensive studies of MIF function have revealed its central rolein innate and adaptive immunity (10). More recently, the abilityof this cytokine to support tumor progression has been highlight-ed, revealing MIF as a potential target for anticancer therapies inmelanomaand colon cancer (40).MIF occurs in immunologicallydistinct conformational isoforms, reduced MIF and oxidized MIF(oxMIF), with the latter predominantly expressed in patients withinflammatory diseases (41) and is highly expressed by variouscancer cell lines (42). This has led to the evaluation of an oxMIF¼blocking antibody (imalumab) in early-phase clinical studies ofselected solid tumors (https://clinicaltrials.gov/ct2/show/NCT01765790). Our findings provide a biological rationale forthe clinical assessment of imalumab or other MIF inhibitors inAML patients.

Activation of PKC signaling pathway has been characterized incancer cells. In hematologic malignancies, different PKC isoformshave been identified as key players in the leukemia microenvi-ronment. In multiple myeloma, pharmacologic inhibition ofactivated PKCbII using enzastaurin inhibited growth factors andcytokines secreted by multiple myeloma–derived bone marrowstromal cells (28). In CLL, PKCb is immediately downstream ofthe B-cell receptor andhas been shown to be important to CLL cellautocrine survival and proliferation in vivo (43). PKCb is alsoessential for the development of CLL in the TCL1 transgenicmouse model, making it a valid therapeutic target in this malig-nancy (44). Furthermore, induction of PKCbII in stromal cells isrequired for the survival of leukemic B cells, and stromal PKCbII isupregulated in samples from CLL, ALL, and MCL patients (28).Our results demonstrate that in primary samples from AMLpatients at diagnosis, PKCb is phosphorylated in the BM-MSCs

in response to MIF stimulation. This leads us to hypothesize thatthis pathway may commonly be activated in other hematologicmalignancies, and moreover, the cancer cell is inducing thisactivation. In summary, this study links secretion of MIF fromprimary AML to a specific BM-MSC pathway, utilizing PKCb tofeedback survival signals, including IL8 secretion toAML. In doingso, we have identified in vitro the potential efficacy of targeting anyone of these molecules to disrupt AML/BM-MSC prosurvivalinteractions.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: A.M. Abdul-Aziz, S.A. Rushworth, K.M. BowlesDevelopment of methodology: A.M. Abdul-Aziz, M.S. Shafat, S.A. Rushworth,K.M. BowlesAcquisition of data (provided animals, acquired and managed patients, pro-vided facilities, etc.): A.M. Abdul-Aziz, M.S. Shafat, M.J. Lawes, S.A. Rushworth,K.M. BowlesAnalysis and interpretation of data (e.g., statistical analysis, biostati-stics, computational analysis): A.M. Abdul-Aziz, T.K. Mehta, F. Di Palma,S.A. Rushworth, K.M. BowlesWriting, review, and/or revision of the manuscript: A.M. Abdul-Aziz,S.A. Rushworth, K.M. BowlesAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): A.M. Abdul-Aziz, M.S. Shafat, S.A. Rushworth,K.M. BowlesStudy supervision: S.A. Rushworth, K.M. Bowles

AcknowledgmentsThe authors wish to thank the Worldwide Cancer Research, The Big C, the

National Institutes for Health Research (United Kingdom), and The Min-istry of Higher Education and Scientific Research of the State of Libya forfunding. We thank Professor Richard Ball and Iain Sheriffs, Norfolk andNorwich tissue bank (United Kingdom) for help with sample collection andstorage.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received April 28, 2016; revised September 22, 2016; accepted October 21,2016; published OnlineFirst November 21, 2016.

References1. Rowe JM, Tallman MS. How I treat acute myeloid leukemia. Blood 2010;

116:3147–56.2. Zeng Z, Shi YX, Samudio IJ, Wang RY, Ling X, Frolova O, et al. Targeting

the leukemia microenvironment by CXCR4 inhibition overcomes resis-tance to kinase inhibitors and chemotherapy in AML. Blood 2009;113:6215–24.

3. Matsunaga T, Takemoto N, Sato T, Takimoto R, Tanaka I, Fujimi A, et al.Interaction between leukemic-cell VLA-4 and stromal fibronectin is adecisive factor for minimal residual disease of acute myelogenous leuke-mia. Nat Med 2003;9:1158–65.

4. Nervi B, Ramirez P, Rettig MP, Uy GL, Holt MS, Ritchey JK, et al. Chemo-sensitization of acute myeloid leukemia (AML) following mobilization bythe CXCR4 antagonist AMD3100. Blood 2009;113:6206–14.

5. Rushworth SA, Murray MY, Zaitseva L, Bowles KM, MacEwan DJ. Identi-fication of Bruton's tyrosine kinase as a therapeutic target in acute myeloidleukemia. Blood 2014;123:1229–38.

6. Lane SW, Scadden DT, Gilliland DG. The leukemic stem cell niche: currentconcepts and therapeutic opportunities. Blood 2009;114:1150–7.

7. Linenberger ML, Jacobson FW, Bennett LG, Broudy VC, Martin FH, Abko-witz JL. Stem cell factor production by humanmarrow stromal fibroblasts.Exp Hematol 1995;23:1104–14.

8. Reuss-Borst MA, Klein G, Waller HD, Muller CA. Differential expression ofadhesion molecules in acute leukemia. Leukemia 1995;9:869–74.

9. Zaitseva L, Murray MY, Shafat MS, Lawes MJ, MacEwan DJ, Bowles KM,et al. Ibrutinib inhibits SDF1/CXCR4 mediated migration in AML. Onco-target 2014;5:9931–9938

10. Nishihira J.Macrophagemigration inhibitory factor (MIF): its essential rolein the immune system and cell growth. J Interferon Cytokine Res 2000;20:751–62.

11. Chen YC, Zhang XW,Niu XH, XinDQ, ZhaoWP, Na YQ, et al. Macrophagemigration inhibitory factor is a direct target of HBP1-mediated transcrip-tional repression that is overexpressed in prostate cancer. Oncogene2010;29:3067–78.

12. Verjans E,Noetzel E, BektasN, Sch€utz AK, LueH, Lennartz B, et al. Dual roleof macrophage migration inhibitory factor (MIF) in human breast cancer.BMC Cancer 2009;9:1–18.

Abdul-Aziz et al.





13. Wilson JM, Coletta PL, Cuthbert RJ, Scott N, MacLennan K, Hawcroft G,et al. Macrophage migration inhibitory factor promotes intestinal tumor-igenesis. Gastroenterology 2005;129:1485–503.

14. Reinart N, Nguyen P-H, Boucas J, Rosen N, Kvasnicka H-M, Heukamp L,et al. Delayed development of chronic lymphocytic leukemia in theabsence of macrophage migration inhibitory factor. Blood 2013;121:812–21.

15. ReikvamH, Fredly H, Kittang AO, BruserudO. The possible diagnostic andprognostic use of systemic chemokine profiles in clinical medicine—theexperience in acute myeloid leukemia from disease development anddiagnosis via conventional chemotherapy to allogeneic stem cell trans-plantation. Toxins 2013;5:336–62.

16. Denizot Y, Fixe P, LiozonE, Praloran V. Serum interleukin-8 (IL-8) and IL-6concentrations in patients with hematologic malignancies. Blood 1996;87:4016–7.

17. Tobler A, Moser B, Dewald B, Geiser T, Studer H, Baggiolini M, et al.Constitutive expression of interleukin-8 and its receptor in humanmyeloidand lymphoid leukemia. Blood 1993;82:2517–25.

18. Schinke C, Giricz O, LiW, Shastri A, Gordon S, Barreyro L, et al. IL8-CXCR2pathway inhibition as a therapeutic strategy against MDS and AML stemcells. Blood 2015;125:3144–52.

19. Rushworth SA, Pillinger G, Abdul-Aziz A, Piddock R, Shafat MS, MurrayMY, et al. Activity of Bruton's tyrosine-kinase inhibitor ibrutinib in patientswith CD117-positive acute myeloid leukaemia: a mechanistic study usingpatient-derived blast cells. Lancet Haematol 2015;2:e204–e11.

20. Falantes JF, Trujillo P, Piruat JI, Calder�on C, M�arquez-Malaver FJ, Martín-Antonio B, et al. Overexpression of GYS1, MIF, andMYC is associated withadverse outcome and poor response to azacitidine in myelodysplasticsyndromes and acute myeloid leukemia. Clin Lymphoma Myeloma Leuk2015;15:236–44.

21. Dios A, Mitchell RA, Aljabari B, Lubetsky J, O'Connor K, Liao H, et al.Inhibition of MIF bioactivity by rational design of pharmacological inhi-bitors of MIF tautomerase activity. J Med Chem 2002;45:2410–6.

22. Leng L, Metz CN, Fang Y, Xu J, Donnelly S, Baugh J, et al. MIF signaltransduction initiated by binding to CD74. J Exp Med 2003;197:1467–76.

23. Bernhagen J, Krohn R, LueH, Gregory JL, Zernecke A, Koenen RR, et al. MIFis a noncognate ligand of CXC chemokine receptors in inflammatory andatherogenic cell recruitment. Nat Med 2007;13:587–96.

24. Park MS, Kim YH, Jung Y, Kim SH, Park JC, Yoon DS, et al. Insiturecruitment of human bone marrow-derived mesenchymal stem cellsusing chemokines for articular cartilage regeneration. Cell Transplant2015;24:1067–83.

25. Liu N, Patzak A, Zhang J. CXCR4-overexpressing bone marrow-derivedmesenchymal stem cells improve repair of acute kidney injury. Am JPhysiol Renal Physiol 2013;305:F1064–73.

26. Barrilleaux BL, Fischer-Valuck BW, Gilliam JK, Phinney DG, O'Connor KC.Activation of CD74 inhibits migration of human mesenchymal stem cells.In Vitro Cell Dev Biol Anim 2010;46:566–72.

27. StarletsD,Gore Y, Binsky I,HaranM,HarpazN, Shvidel L, et al. Cell-surfaceCD74 initiates a signaling cascade leading to cell proliferation and survival.Blood 2006;107:4807–16.

28. Lutzny G, Kocher T, Schmidt-Supprian M, Rudelius M, Klein-Hitpass L,Finch AJ, et al. Protein kinase c-beta-dependent activation of NF-kappaB instromal cells is indispensable for the survival of chronic lymphocyticleukemia B cells in vivo. Cancer Cell 2013;23:77–92.

29. JuliussonG, Antunovic P, Derolf A, Lehmann S,Mollgard L, StockelbergD,et al. Age and acute myeloid leukemia: real world data on decision to treatand outcomes from the Swedish acute leukemia registry. Blood 2009;113:4179–87.

30. Burnett A.Ham-Wasserman Lecture. Treatment of acutemyeloid leukemia:are we making progress? Hematology 20122012:1–6

31. Bruserud�, Ryningen A, Olsnes AM, Stordrange L,�yan AM, Kalland KH,et al. Subclassification of patients with acute myelogenous leukemia basedon chemokine responsiveness and constitutive chemokine release by theirleukemic cells. Haematologica 2007;92:332–41.

32. Reikvam H, Brenner AK, Hagen KM, Liseth K, Skrede S, Hatfield KJ, et al.The cytokine-mediated crosstalk between primary human acute myeloidcells and mesenchymal stem cells alters the local cytokine network andthe global gene expression profile of the mesenchymal cells. Stem Cell Res2015;15:530–41.

33. Brenner AK, Reikvam H, Bruserud �. A subset of patients with acutemyeloid leukemia has leukemia cells characterized by chemokine respon-siveness and altered expression of transcriptional as well as angiogenicregulators. Front Immunol 2016;7:205.

34. Huang JC, Basu SK, Zhao X, Chien S, Fang M, Oehler VG. Mesenchymalstromal cells derived from acute myeloid leukemia bone marrow exhibitaberrant cytogenetics and cytokine elaboration. Blood Cancer J 2015;5:e302

35. Hartl D, Latzin P, Hordijk P, Marcos V, Rudolph C, Woischnik M, et al.Cleavage of CXCR1 on neutrophils disables bacterial killing in cysticfibrosis lung disease. Nat Med 2007;13:1423–30.

36. Gao Y, Guan Z, Chen J, Xie H, Yang Z, Fan J, et al. CXCL5/CXCR2 axispromotes bladder cancer cell migration and invasion by activatingPI3K/AKT-induced upregulation of MMP2/MMP9. Int J Oncol 2015;47:690–700.

37. Nomellini V, Brubaker AL, Mahbub S, Palmer JL, Gomez CR, Kovacs EJ.Dysregulation of neutrophil CXCR2 and pulmonary endothelial ICAM-1 promotes age-related pulmonary inflammation. Aging Dis 2012;3:234–47.

38. Inoue K, Slaton JW, Eve BY, Kim SJ, Perrotte P, BalbayMD, et al. Interleukin8 expression regulates tumorigenicity and metastases in androgen-inde-pendent prostate cancer. Clin Cancer Res 2000;6:2104–19.

39. Fu XT, Dai Z, Song K, Zhang ZJ, Zhou ZJ, Zhou SL, et al. Macrophage-secreted IL-8 induces epithelial-mesenchymal transition in hepatocellularcarcinoma cells by activating the JAK2/STAT3/Snail pathway. Int J Oncol2015;46:587–96.

40. Ioannou K, Cheng KF, Crichlow GV, Birmpilis AI, Lolis EJ, TsitsilonisOE, et al. ISO-66, a novel inhibitor of macrophage migration, showsefficacy in melanoma and colon cancer models. Int J Oncol 2014;45:1457–68.

41. Thiele M, Kerschbaumer RJ, Tam FW, Volkel D, Douillard P, Schinagl A,et al. Selective targeting of a disease-related conformational isoform ofmacrophage migration inhibitory factor ameliorates inflammatory condi-tions. J Immunol 2015;195:2343–52.

42. Schinagl A, Hagemann T, Douillard P, Thiele M, Voelkel D, Freissmuth M,et al. Abstract 4841: Oxidized macrophage migration inhibitory factor(oxMIF) expressed by tumor stroma and tumor cells, contributes to tumorgrowth. In: Proceedings of the 105th Annual Meeting of the AmericanAssociation for Cancer Research; 2014 Apr 5–9; San Diego, CA. Philadel-phia, PA: AACR. Abstract nr 4841.

43. El-Gamal D, Williams K, LaFollette TD, Cannon M, Blachly JS, Zhong Y,et al. PKC-beta as a therapeutic target in CLL: PKC inhibitor AEB071demonstrates preclinical activity in CLL. Blood 2014;124:1481–91.

44. Holler C, Pinon JD, Denk U, Heyder C, Hofbauer S, Greil R, et al.PKCbeta is essential for the development of chronic lymphocyticleukemia in the TCL1 transgenic mouse model: validation of PKCbetaas a therapeutic target in chronic lymphocytic leukemia. Blood2009;113:2791–4.






Published OnlineFirst November 21, 2016.Cancer Res Amina M. Abdul-Aziz, Manar S. Shafat, Tarang K. Mehta, et al. Myeloid Leukemia

/IL8 Is Essential in Human AcuteβMIF-Induced Stromal PKC

Updated version

10.1158/0008-5472.CAN-16-1095doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2016/11/19/0008-5472.CAN-16-1095.DC1

Access the most recent supplemental material at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://cancerres.aacrjournals.org/content/early/2017/01/09/0008-5472.CAN-16-1095To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-16-1095

http://cancerres.aacrjournals.org/content/suppl/2016/11/19/0008-5472.CAN-16-1095.DC1

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/early/2017/01/09/0008-5472.CAN-16-1095


mif-induced stromal pkcβ/il8 is essential in human acute ... · microenvironment and immunology...

Documents